Design of Machine Learning-Based Software for Transportation Optimization and Accident Reduction

Arce-Cocha G.; Romero-Berrospi D.; Avalos-Levano C.; Ramos-Cosi S. · International Journal of Engineering Trends and Technology (2025) · DOI: 10.14445/22315381/ijett-v73i7p107

Resumen

Traffic congestion and accidents have been critical problems in urban environments, especially in cities with limited infrastructure and a lack of updated data. This work presented the conceptual design of an application aimed at optimizing routes, improving urban mobility, and contributing to the reduction of road incidents. The methodology used was Design Thinking, which allowed for the identification of user needs and the proposal of innovative solutions through the integration of technologies such as Flutter, Firebase, and TensorFlow Lite. The proposed design included algorithms capable of analyzing real-time data from sensors, cameras, and GPS devices, with the goal of anticipating traffic patterns and generating more efficient routes. Additionally, functionalities such as a scoring system to evaluate driver behavior and an administrative dashboard for managing alerts and traffic analysis were proposed. It was concluded that the conceptual design exhibited significant potential to improve road safety in high-congestion areas and promote more efficient mobility. Future research suggested the development or implementation of this design, in addition to expanding the system’s geographic coverage and evaluating its impact on the sustainability of urban transport.

Instituciones: Universidad de Ciencias y Humanidades